• Årsmøde
• Tidligere møder og ekskursioner

Årsmødet og Generalforsamling afholdes på Geocenter København, Øster Voldgade 10 lørdag den 6. marts 2010 med temaet:  ”Geologisk modellering – Muligheder og begrænsninger” Årsmødet fokuserer i 2010 på computerbaseret modellering inden for de geologiske discipliner. Matematisk modellering har vundet indpas inden for mange aspekter af geologien fra magmatisk udvikling i grænsen mellem skorpen og kappen, udbreddelse af seismike bølger og refleksionsseismisk analyse, temperaturudviklingen af sedimentære bassiner og modning af kildebjergarter, over olie og gas reservoirer, grundvandsmagasiner til palæoklima. Det er intentionen med mødet ikke bare at afdække nogle af de mange anvendelser af modellering, men også at få belyst det potentiale der ligger i dette værktøj samt de begrænsninger og udfordringer det nuværende arbejde har afdækket. efterfulgt af Generalforsamling og festmiddag med uddeling af Danmarks Geologipris 2009.

Mødet sponsoreres af:

Tilmelding senest tirsdag den 4. marts 2010 kl. 10 (se nedenfor). Der vil kunne opnås støtte til rejseudgifter for medlemmer vest for Storebælt (se nedenfor).

Program

Rækkefølgen af titler er efter emne og afspejler ikke rækkefølgen af indlæg i foredragsrækken

Global warming and mass extinctions caused by sediment degassing of volcanic basins with emphasis on the Paleocene-Eocene Thermal Maximum (PETM)
Sverre Planke (Volcanic Basins Petroleum Resarch, Oslo) og Henrik Svensen (Oslo Universitet).

Modelling in Earth system science – climate, carbon and sedimentation on geological timescale as influenced by biological activity
Christian J. Bjerrum, IGG

Modellering af fin-skala jordstrukturer: Hvordan opnår vi detaljeret viden om jordstrukturer fra kappeniveau til nær jordoverfladen?
Lars Nielsen & Hans Thybo (IGG)

Modelling the geochemistry of high-temperature magmatic processes
Tod Waight (IGG)

Geologisk modellering i forbindelse med den nationale grundvandskortlægning 
Jette Sørensen (Rambøll)

Geologisk modellering med GeoScene3D
Niels Peter Jensen eller Lars Nebel ( I-GIS)

Geologisk modellering af Københavns undergrund i forbindelse med forundersøgelserne til Cityringen
Henrik Olsen, Sanne L. Hansen, Thomas B. Andersen og Ole F. Nielsen (COWI AS)

Reservoirmodellering til CO2-lagring
Peter Frykman (GEUS)

Modelling diagenetic phenomena in a sandstone reservoir – ideas and results from Siri Main Field, North Sea
Johan Svendsen & Mette Lise Kjær Poulsen (DONG Energy)

Modelling a heavily remobilised, deep marine sandstone reservoir – ideas and results from Nini Main Field, North Sea
Christian Mogensen, Tine Holmer & Johan Svendsen (DONG Energy)

Seismic modelling and inversion of a Paleocene submarine fan system
Nur Ucisik, Svein Idar Førsund (Noreco) and Marc Beller (Mærsk Olie og Gas AS)

From rocks to reservoir models: the rock physics !
Igor Escobar (Mærsk Olie og Gas AS)

Sammendrag

Global warming and mass extinctions caused by sediment degassing of volcanic basins with emphasis on the Paleocene-Eocene Thermal Maximum (PETM)

Sverre Planke( 1,2) og Henrik Svensen (2)

(1) Volcanic Basin Petroleum Research, Oslo, Norway – contact Sverre Planke
(2) Physics of Geological Processes, University of Oslo, Norway

We have for the past decade worked on a hypothesis linking major environmental changes in Earth history to the eruption of Large Igneous Provinces (LIPs). It is well known that there are temporal correlations between LIPs and periods of global warming and/or mass extinctions. However, there are several problems explaining the environmental changes by extrusive volcanism. In particular, (1) there is no correlation between the size of a LIP and the environmental impact, (2) it is not possible to explain observed increases in isotopically light carbon by degassing of mafic magma, and (3) it is not sufficient carbon in mafic magma to explain the mass of atmospheric carbon needed to trigger global warming.

Our hypothesis, initially published in 2004, stresses the emplacement environment. Sedimentary basins contain large volumes of volatiles or rocks that can be devolatized by heating. Magma intruding sedimentary basins will heat the host rock, causing massive degassing of the contact metamorphic aureoles. The type volatiles that are formed in the aureoles are strongly dependent on the host rock lithologies. Greenhouse gases, such as CH4 and CO2, are formed when organic-rich rocks or carbonates are heated, whereas poisonous gases (SO2, halocarbons, etc.) are formed when evaporate-rich sequences are heated.

In contrast, only boiling or formation of supercritical water is taking place when barren sandstones or shales are heated. The sediment degassing hypothesis has been tested by focused studies of (1) the Paleocene-Eocene Thermal Maximum (PETM) and the Northeast Atlantic Igneous Province, (2) the Toarcian Oceanic Anoxic Event and the Karoo LIP, and (3) the end-Permian extinction and the Siberian Traps LIP. Borehole studies of aureole rocks in all three igneous provinces reveal a very high production potential of greenhouse gases. Thousands of kilometer-sized vent complexes were further formed in all three provinces, documenting focused fluid flow from the aureoles to the atmosphere.

The causal link between the intrusive volcanism and the environmental changes has further been strengthened by high-precision U/Pb dating of the sill complexes. The new dates show that the age of the intrusive event overlap with the age of the environmental changes. Heating experiments of petroleum-bearing evaporates from the Tunguska Basin in Siberia document that ozone-destroying halocarbons may have formed in large quantities during the end-Permian, and show that the host rock lithogy is a critical factor when assessing the environmental implications of intrusive magmatism.

Modellering af fin-skala jordstrukturer: Hvordan opnår vi detaljeret viden om jordstrukturer fra kappeniveau til nær jordoverfladen?

Lars Nielsen & Hans Thybo (IGG)

Seismiske bølgefelter har i udstrakt grad været brugt til at uddrive information om jordstrukturer på global, regional og lokal skala. Radarbølger bruges i større og større udstrækning til at kortlægge de øverste meter af jorden på lokal skala.

Der findes mange forskellige metoder til at modellere sådanne bølgefelter. I denne præsentation gives eksempler på, hvordan modellering af seismiske bølgefelter og radarbølger har givet ny viden om finskala strukturer fra jordens kappe til nær jordens overflade.

Modelling the geochemistry of high-temperature magmatic processes

Tod Waight (IGG)

Modern geochemical analytical techniques have revolutionised our ability to model the compositions of melts generated or modified during partial melting, fractional crystallisation and contamination.

Interactions between intruded magmas and their host country rocks provides a good case example, where the models used have evolved with time as our understanding of these processes have also evolved as a consequence of experimental, theoretical and field-based observations. The presentation will include a general overview of the processes, theory and assumptions involved, together with potential pitfalls involved.

Modelling in Earth system science – climate, carbon and sedimentation on geological time scale as influenced by biological activity

Christian J. Bjerrum, IGG

Modelling of the carbon cycle and climate on geological time scales is still in its infancy, despite a nearly complete semi-quantative treatment of the carbon-oxygen cycles by Ebelmen already in 1845.

Simultaneously, Fourier (1827), Tyndall (1861) and Arrhenius (1896) identified and determined the long wave absorption by CO2 and its possible climate impact. Through modelling I will illustrate how biological evolution and activity most likely also have had a major influence on climate and element cycling – shaping Earth evolution. The modelling highlights problems with current paradigms on carbon cycling and generates new hypotheses to be tested by fundamental field work.

Geologisk modellering i forbindelse med den nationale grundvandskortlægning

Jette Sørensen (Rambøll)

Foredraget vil omhandle mål, fremgangsmåder og erfaringer indhøstet i forbindelse med opstilling af geologiske modeller ved Miljøcentrenes grundvandskortlægning. Der vil både blive præsenteret egne eksempler og eksempler fra andre projekter som bl.a. GEUS afdeling for grundvandskortlægning har været involveret i.

Geologisk modellering med GeoScene3

Niels Peter Jensen eller Lars Nebel ( I-GIS)

I-GIS vil præsentere firmaets IT-værktøj til geologisk 3D modellering og visualisering, og bl.a. komme ind på hvilke data man kan inddrage i opstillingen af modeller. Programmet er udviklet i samarbejde med Miljøcentrene til modellering ved grundvandskortlægning, men vil i høj grad også kunne anvendes ved andre typer af geologisk 3D modellering og visualisering.

Geologisk modellering af Københavns undergrund i forbindelse med forundersøgelserne til Cityringen

Henrik Olsen, Sanne L. Hansen, Thomas B. Andersen og Ole F. Nielsen (COWI AS)

I de kommende år vil den eksisterende Metro i København blive suppleret med Cityringen, en 16,3 km lang ringbane, der vil forbinde det indre København med Østerbro, Nørrebro og Frederiksberg.

Forundersøgelserne er netop afsluttet, og som en vigtig del af disse undersøgelser er der opstillet en 3D digital geologisk model over Københavns undergrund. Modellen er baseret på en integreret analyse af mange datatyper. Der indgår mere end 5000 boringer, hvoraf ca. 500 er udført i forbindelse med forundersøgelserne. Hertil kommer logstratigrafisk analyse af geofysiske logs fra mere end 300 boringer samt refraktionsseismisk tomografi og refleksionsseismisk tolkning af 13 km ny seismik skudt langs Københavns gader og veje.

Den geologiske model giver et detaljeret billede af de kvartære aflejringers komplicerede opbygning og afslører et system af begravede dale, der skærer sig dybt ned i prækvartæroverfladen. Modellen indeholder også nye oplysninger  om folder og forkastninger i Københavnkalken (Danien) og giver et detaljeret indblik i udbredelsen af Lellinge Grønsand (Selandien).

Modellen skal sammenholdes med den planlagte tunnel og anvendes til tunnelentreprenørernes vurdering af gravbarhed. Den skal også anvendes til vurdering af de geotekniske forhold og grundvandsforholdene, blandt andet i forbindelse med etableringen af de 17 undergrundsstationer langs Cityringen.

Reservoirmodellering til CO2-lagring

Peter Frykman (GEUS)

Reservoirmodeller til CO2-lagre stiller nye krav til udformning og skala, da det delvist er andre processer der skal studeres, end hvad der er normalt for olie/gas reservoirer. Især trykopbygning og risiko-analyse udfordrer modelløren.

Den geologiske model har stor betydning for kapaciteten og opførslen af lageret under fyldning, og den store mobilitet af den superkritiske CO2 giver anledning til store effekter fra heterogeniteter.

Modelling diagenetic phenomena in a sandstone reservoir – ideas and results from Siri Main Field, North Sea

Johan Svendsen & Mette Lise Kjær Poulsen

Understanding diagenetic processes in the reservoir is often of crucial importance for an optimal development of an oil-field. But implementing results and ideas obtained at micron scale into a geomodel with cells of 100 metres size is a challenge.

In the Siri Field, Danish North Sea, a number of diagenetic processes are recognized as being important for understanding and developing the field. Isotopic data from a number of carbonates suggests that the cementation was formed at the OWC. A

As the field has experienced multiple pulses of oil-inflow and/or structural tilt, a number of OWCs have been present in the field over time, and a number of cemented horizons are therefore seen. Furthermore, the oil-fill impacts the evolution of the diagentic clay, which results in different permeability above and below the OWC. These phenomena, and there subsequent geocellular modelling is the scope for this presentation.

Modelling a heavily remobilised, deep marine sandstone reservoir – ideas and results from Nini Main Field, North Sea

Christian Mogensen, Tine Holmer & Johan Svendsen (DONG Energy)

The Nini Main Field, Danish North Sea, is characterized by intense remobilization. Sands, although in full dynamic communication, are placed at varying stratigraphic levels. As the sands are close to seismic resolution, depicting these remobilized phenomena on seismic is often difficult.

A new Simultaneous inversion has provided an excellent tool for mapping the sand, and proving the stratigraphically cross-cutting nature. Furthermore, reservoir properties are dependent on the degree of remobilisation. A detailed geological concept, as well as full usage of core data and seismic inversion, has allowed for a more detailed modelling of the Nini Main Field.

Seismic modelling and inversion of a Paleocene submarine fan system

Nur Ucisik, Svein Idar Førsund (Noreco) and Marc Beller (Mærsk Olie og Gas AS)
A seismic modelling study has been performed in order to better understand the causes of observed amplitude increase in the seismic data from a sub-marine fan system of Paleocene age. The modelling was conducted for a wide range of variations in porosity, water saturation and reservoir thickness, as well as removing various lithological units from the well data used.

The synthetic offset modelling and wedge modelling results suggest that the observed far-offset amplitude increase is not related to a hydrocarbon effect at the Top Borr reflection.

The apparent AVO effect is neither influenced by lithologies in the lower Borr to Chalk nor the high velocity zone in the Borr sand itself. The increase in amplitude at far offsets is related to interference from the Chalk reflection. The Top Chalk reflection is close to the critical angle at 2750 m offset and the critical angle is reached at 3250 m.

The 3D seismic dataset has also been inverted for relative P-impedance. The aim of the inversion was to improve the interpretability of the seismic dataset with focus on top reservoir (Top Borr sand).

The inversion results show that the P-impedance data improves both the interpretability of Top Borr sand and the high velocity zone above the Borr sand.

From rocks to reservoir models: the rock physics!

Igor Escobar (Mærsk Olie og Gas)

In hydrocarbon exploration and production, establishing a reliable model of the subsurface is paramount. Typically these models include spatial distributions of both static and dynamic physical/chemical properties describing the rocks and fluids.

Estimating these models is an exercise that is far from trivial since it involves combining data of different types, different resolutions, and different spatial coverage. Commonly we have to deal with rock and fluid samples taken along wells, wireline data, surface and borehole seismic data, and geological conceptual models among others.

Along the whole process, theoretical and empirical physical and chemical models of the rocks and fluids need to be established as a mean to link the different data types. Also statistical/geostatistical techniques together with inversion methods are commonly used as a way to overcome the differences in resolution and as part of the data assimilation process.

During the presentation I would address some of these issues, especially those related to the understanding of the rock physics and the estimation of static reservoir properties. It is intended to be an overview of best practices today and the way forward towards the definition of shared earth models.

Posterabstracts

Distinguishing the preboreal oscillations using tephrostratigraphy

Jeppe Joel Larsen (IGG)

Commonly only one Preboreal oscillation is recorded in terrestrial and lacustrine sedimentary successions. The climatic appearance of this oscillation differs from different North European localities, and is thus not always interpreted to be the same. In some sedimentary succession two climatic changes are observed, and in even fewer three.

Because of two contemporaneous 14C-plateaus, it is very difficult to precisely date the events, therefore making vital chronostratigraphic correlation complicated. Here I use the Hässeldalen Tephra to precisely identify the Preboreal Oscillation (PBO) as defined by Björck et al. (1997). This tephra horizon is an important isochron for future research on the Preboreal oscillations.

Shelf-ocean exchange with different idealized shelf profiles: Preliminary results for a coastal upwelling system

Muchamad Al Azhar & Crhistina J. Bjerrum (IGG)

The continental margin is considered as a significant area for oceanic biogeochemical cycles. It receives abundant nutrients from the deep-open ocean through ocean-shelf exchange, especially coastal upwelling. Understanding upwelling variability as function of stratification and shelf bathymetry provide insight on regulating coastal ecosystems.

Coastal upwelling for narrow (California-like) and wide (Canary Current-like) idealized shelf profiles are investigated using the Regional Ocean Modeling System (ROMS) under simplified boundary condition and forcing. Present day simplified boundary conditions are used to reproduce and validate the models fit to first order observed spatial mean variation.

Such ‘degraded’ boundary condition will then be used to investigate scenarios of geological past condition with implication for climate change. Simulation results indicate that shelf width control exchange between deep ocean and shelf area by its influence on mixing, mesoscale eddy activity, and frontal formation.

Cenozoic uplift and subsidence in the North Atlantic region: Geological evidence revisited

Ingrid Anell,  H. Thybo & I. Artemieva (IGG)

The topographic evolution of the “passive” margins of the North Atlantic during the last 65 My is the subject of extensive debate. The various geological, geomorphological and geophysical methods for studies of uplift and subsidence have each their limitations.

We have compiled a database of studies of topographic changes in the North Atlantic region during the Cenozoic (65-0 Ma), based on published results, with focus on sign, time and amplitude (where possible) of these changes. Our compilation, based on reflection seismic studies, AFTA (apatite fission track analysis), VR (vitrinite reflectance) trends, maximum burial, sediment supply studies, mass balance calculations and extrapolation of seismic profiles to onshore geomorphological features, indicates that a first major phase of regional uplift occurred in the late Paleocene-early Eocene (c. 60-50 Ma). This uplift was probably related to the break-up of the Atlantic between Europe and Greenland, since the timing of uplift propagated northward.

The second major phase of regional uplift that affected all marginal areas of the North Atlantic occurred in the Plio-Pleistocene (5-0 Ma). The importance of a tectonic component for this uplift is inconclusive but erosion-driven glacio-isostatic compensation is likely to have been an important factor.

Thermobarometriske modellering af klinopyroxener fra Argentinske basalter

Simon Normann Lauritsen & Paul Martin Holm (IGG)

I relation til back-arc vulkanismen i det vestlige Argentina forsøges det med thermobarometriske beregninger, at bestemme tryk og temperatur for dannelse af klinopyroxener. Denne beregning foretages ved brug af fire modeller, hvor andelen af jadeit bestemmes i cpx. Til bestemmelse af jadeit komponenten antages det, at den mindste mængde af Na eller VIAl, bestemmer mængden af jadeit, og derved beliggenheden af et eller flere magmakamre.

Isotopic and elemental evidence of Icelandic plume involvement during late Cretaceous initial rifting of the Fram Strait gateway and generation of the Kap Washington Group basalts (NE Greenland)

Helene I. Duprat, P.M. Holm, S.B. Thorarinsson , C. Tegner & R. Harlou (IGG)

The Kap Washington Group (KWG) Province in North Greenland consists of a suite of alkaline rhyolites and basalts extruded ~ 70 – 61 Ma (Þórarinsson et al., (2009) and occur on the northern tips of Lockwood Ø, Kap Kane, Kap Washington and Kap Cannon. The volcanic province is a result of late Cretaceous shearing processes related to the initial separation of the North American Plate and the Eurasian Plate. Some 33 Ma ago the sheared margin rifted and broke up obliquely and
thus connected the most northerly part of the Mid Atlantic Ridge, the Knipovich spreading ridge, via Lena Through, with the Gakkel spreading ridge in the Arctic (Engen et al., 2008).

The most radiogenic KWG samples define arrays in the 208Pb/204Pb – 206Pb/ 204Pb and 207Pb/204Pb – 206Pb/ 204Pb diagrams that run parallel to and above that of the “Iceland plume trend”. High 208Pb/204Pb compositions are also seen in lavas from the western part of Gakkel Ridge (Goldstein et, fresh pillow basalts from Lena Trough (Nauret et al., J. Petrol., in press) and the most radiogenic samples from Knipovich Ridge (Blichert-Toft et al., 2005).

Lherzolitic xenoliths from volcanic centres in NW Spitsbergen, show similar high208Pb/204Pb (Ionov et al., 2002a & b), indicating that this special isotopic feature of the region could be a lithospheric signal. However, high208Pb/204Pb compositions are also seen in basalts and rhyolites of the Vestbrona Formation which is highly silica-undersaturated alkaline volcanic rocks (Prestvik et al., 1999) of early paleogene age (~55 Ma, (Bugge et al., 1980)) located beyond the eastern terminations of the Jan Mayen Fracture-zone on the Norwegian continental shelf.

The basalts of the Vestbrona are likely to represent melts that ascended via old lithospheric faults from the mantle below Jan Mayen and could record an enriched end member component of the proto-Icelandic plume (Torske and Prestvik 1991). Isotopic and elemental modeling suggest that these old Vestbrona rocks and the KWG group 2 basalts share a common mantle end member with high Th/Nb and radiogenic lead. This could indicate either a compositionally homogenous lithosphere in the region, or that the KWG basalts are a result of an early Icelandic Plume activity

Geochronology af et vulkansk område i det sydlige Mendoza, Argentina.

Jónas Gudnason (IGG)

Den kvartære vulkanisme i et mindre område i det sydlige Mendoza i Argentina er afgrænset af nye 40Ar/39Ar dateringer. Det undersøgte område er opbygget af lavaer fra tre stærkt sammensvejsede skoriekegler.

Keglernes relative alder har været afgrænset ud fra feltrelationer og daterede prøver valgt til at præsenterer de ældste og yngste lavaer. Vores analyser viser at området har været aktivt fra Pleistocæn til Holocæn med mulighed for endnu yngre vulkanitter.

Geokemiske undersøgelser af Infernillo-komplekset i back-arc provisen Paynia, sydvestlige Medoza, Argentina

Mia Nielsen (IGG)

I Andesbjergenes udløbere i den vestlige udkant af Malargüe ”fold and thrust”-bæltet ligger Infernillo-komplekset. Komplekset udgør den nordvestlige udkant af den vulkanske provins Payenia, som er resulteret af den Miocæne til recente back-arc spredning i provinsen.

Her er to meget unge basaltiske ’a’a-flows (Media Luna og Infernillo) blevet studeret og prøvetaget, og på den baggrund er en eruptionshistorie forsøgt etableret. I områdets nærhed ligger flere andre lavaer, fra hvilke der også blev indsamlet prøver. Disse lavaflows synes at være af varierende alder, nogle lige så unge som Infernillo-komplekset og mindst 2 synes ældre bedømt ud fra graden af erosion. Tilsammen udgør disse lavaer et isoleret vulkansk område med mindst 11 forskellige udbrudssteder.

Klima- og havniveausvingninger fra den mellemste og sidste del af Holocæn, tolket på basis af geokemi og sedimentologi i en borekerne på 14 meter fra Klintsø, NV Sjælland, DK.

Rikke Jørgensen

Der er blevet boret en 14 meter sedimentkerne i Klintsø, NV Sjælland, med det formål at etablere tidspunkterne for de fire littorinatransgressioner, der ses i det nærliggende Tengslemark bassin og især afgøre, om der finder en femte sen-subboreal transgression sted på Sjælland, som det ses i havniveaukurverne for den nordlige del af Nordsøen (Behre, 2007).

Der er blevet lavet en sedimentologisk tolkning på kernen, samt geokemi. Der er målt TC, TS, CaCO3, TOC og magnetisk susceptibilitet, som bruges som klimaproxy. Derudover er der sendt 8 prøver til datering. Ud fra data indsamlet i de nærliggende Tengslemark og Højby sø kan havniveauændringer og til dels klimaet estimeres, hvilket er det endelige formål med dette speciale.

Crustal structure of the Oslo Graben: preliminary results from MAGNUS-REX, new crustal scale refraction profiling across southern Norway.

Wanda Stratford, H. Thybo, J. Faleide, O. Olesen & A. Tryggvason

A new crustal scale seismic profile was recorded in October, 2007, across the Oslo Graben as part of the Magnus-Rex seismic project. The Oslo line passes through the middle of the graben, through the region of lowest gravity anomaly and across the lowest gravity gradients at the margins. Single component seismographs were deployed along the line at 2 km spacing, except for a 120 km wide section across the graben where the instrument spacing was reduced to 750 metres. Seven shots of 100-400 kg charge were fired along the Oslo line.

Preliminary results from the Oslo line show a Moho depth of around 36-38 km in western Norway. The Moho is interpreted as shallow only by ~1-4 km beneath the Oslo Graben.

Cisternerne under Søndermarken

Christian Brogaard Pedersen & Katrine Bruhn Holck (IGG)

Inden for de sidste få år har udforskningen af gamle historiske bygninger og kunstige huler, også kaldet ’urban eksploration’, såsom gamle vandværker, undergrundspassager, broer, kloakker o.l. fået en stigende interesse. Disse områder rummer ofte kalkudfældninger, der morfologisk set minder meget om naturlige drypsten (speleotemer). Interessen for disse forekomster af urbane drypsten har de seneste par år vakt øget undren og fascination. Hvorfor er de der, hvad består de af, hvordan dannes de og har de nogen relation til naturlige speleotemer?

De urbane drypstens morfologi afspejler flere genkendelige strukturer, der tilnærmelsesvis kan klassificeres under de samme drypstenstyper som de naturlige forekomster. De har så mange karakteristika tilfælles med speleotemer, at de næsten kan forveksles med hinanden, men kan alligevel ikke betragtes som ’egentlige’ speleotemer, idet de ikke er dannet i en ’naturlig’ hule. Alligevel betegner vi dem som drypsten.

Nærstudier blev udført på disse urbane drypsten for at undersøge om hvorvidt de indeholder proxier, der kan anvendes som et klimaarkiv, ved at studere deres isotopiske, kemiske og morfologiske forhold i Cisternerne under Søndermarken.